Generative AI Weekly - May 2, 2024

This week:

• Braintrust launches AIR - "the first AI Recruiter." It handles job description development, posting, scheduling/interviewing, scoring, and recommendations. "We are transforming hiring to remove 95% of the time, cost, and effort to hire top talent." I've seen the demo, and it's quite something. If I could use it today, I would!
• Morning Consult and Canva released a Marketing & AI Report. It is a survey of over 4,000 marketing and creative leaders. Key insights include the complexity of keeping up with new technological advancements, the lack of corporate policy, and the potential for time-saving applications.
• Digiday's recent AI Trends Report for E-Commerce Marketers (co-sponsored by Klaviyo and Qualtrics). Generally, I find DTC marketers to be early adopters of martech across the board, and this report hints that we need to think beyond just Generative AI as a potential tool. There are some nice use cases from actual brands, which are certainly inspiring!
• The bi-annual CMO survey is always worth a read, but this spring's Top Line Report provides industry benchmarks for AI adoption. Current mean estimations point to 5-7% improvements in sales productivity, marketing overhead costs, and customer satisfaction.
• A David Kellogg presentation on AI's Impact on Go-To-Market. Key takeaways are around coming board-level productivity expectations - whether ahead of the curve or not.
• Ethan Mollick is profiled by the Wall Street Journal. He recently released a "framework finder" GPT which I love.
• To that end, NBER on The Economics of Generative AI. "On average, worker productivity increased by 14 percent…"
• A fun article from March 2023 on how software is set to change structurally as Gen AI gets adopted in ways we might not even notice.
• Two excellent articles on what LLMs are and are not good at.
• The Gradient is a group of Standford AI Lab students that offers a balanced view of AI adoption. Their recent article, "Financial Market Applications of LLMs" is a nice example of their editorial style. 
• Our World in Data - a favorite research institution, has some great datasets on AI investment and performance globally.
• Github teases its new Copilot Workspace - "where any developer can go from idea, to code, to software all in natural language."
• And for a paper of the week… "ChatGPT 'contamination': estimating the prevalence of LLMs in the scholarly literature" hinting towards some potent, fresh, and ingenious tricks to identify GPT written work.